---
title: "Discussion on \"From Ingress to Inference Gateway: Gateway API and the Inference Extension\" | Hashnode"
description: "Discussion on \"From Ingress to Inference Gateway: Gateway API and the Inference Extension\". How Kubernetes networking became model-aware: the Gateway API resource model, the Inference Extension's InferencePool, and the endpoint picker that routes on KV-cache and queue metrics instead of round-robin."
url: "https://hashnode.com/posts/from-ingress-to-inference-gateway-gateway-api-and-the-inference-extension/6a8fd785ca3ff895ff867ab8"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [From Ingress to Inference Gateway: Gateway API and the Inference Extension](https://todea.hashnode.dev/from-ingress-to-inference-gateway-gateway-api-and-the-inference-extension?utm_source=hashnode&utm_medium=feed)

How Kubernetes networking became model-aware: the Gateway API resource model, the Inference Extension's InferencePool, and the endpoint picker that routes on KV-cache and queue metrics instead of round-robin.

Aug 27 · 19 min read · todea.hashnode.dev  
Read the article

-   [#kubernetes](https://hashnode.com/tag/kubernetes)
-   [#gatewayapi](https://hashnode.com/tag/gatewayapi)
-   [#vllm](https://hashnode.com/tag/vllm)
-   [#llm](https://hashnode.com/tag/llm)
-   [#aiinfrastructure](https://hashnode.com/tag/aiinfrastructure)
-   [#inference](https://hashnode.com/tag/inference)
-   [#cloudnative](https://hashnode.com/tag/cloudnative)

## Written by

[Ivan Porta](https://hashnode.com/@gtrekter)

@gtrekter · Seoul, South Korea

  
[View profile](https://hashnode.com/@gtrekter)

Founder @ Todea

Founder @ Todea | Cloud-native platform engineering for production.

## Responses

Markdown supported · ⌘↵ to post

No responses yet.
